22 May 2025 in International Journal of Pattern Recognition and Artificial Intelligence
25 July 2025 in Journal of Hydroinformatics
Detection of submerged objects in underwater environments is uniquely challenging due to low visibility, variable light conditions, and high noise and distortion in Underwater Aquatic System (UAS). The traditional object detection models, including the most recent versions of You Only Look Once (YOLO), drop too often in accuracy and robustness under such hostile conditions. To handle these, YOLO-Transformer Hybrid model is proposed which combines YOLO's real-time detection capabilities with enhanced contextual understanding accorded by transformer-based attention mechanisms. The YOLO backbone extracts the essential features from an input image and processes the image through a number of convolutional layers for real-time object detection. It adds attention layers to the transformer module, focusing on relevant regions of the image to extract long-range dependencies and contextual relationships possibly missed by convolutional layers in isolation. Then augment this hybrid approach with Internet of Things (IoT) sensor data to add valued-added environmental context that may improve detection accuracy. IoT sensors can facilitate dynamic adaptation to the variations in underwater conditions by continuously providing real-time data on temperature, turbidity, etc. The proposed model ensures 97.5% detection accuracy, outperforming traditional YOLO models under challenging underwater scenes.
14 September 2023 in International Journal of Interactive Mobile Technologies (iJIM)
IoT is becoming increasingly popular due to its quick expansion and variety of applications. In addition, 5G technology helps with communication and network connectivity. This work integrates C-RAN with IoT networks to provide an experimental 5G testbed. In a 5G IoT environment, this experience is utilized to enhance both perpendicular and flat localization (3D localization). DRCaG, an acronym for the proposed model, stands for a deep, complicated network with a gated layer on top. The performance of the proposed model has been demonstrated through extensive simulations in terms of learning reduction, accuracy, and matrix disorientation, with a variable signal-to-noise ratio (SNR) spanning from 20 dB to + 20 dB, which illustrates the superiority of DRCaG compared to others. An online, end-to-end solution based on deep learning techniques is presented in this study for the fast, precise, reliable, and automatic detection of diverse petty crime types. By detecting tiny crimes like hostility, bag snatching, and vandalism, the suggested system may not only identify unusual passenger behavior like vandalism and accidents but also improve passenger security. The solution performs admirably in a variety of use cases and environmental settings.
10 October 2025 in Surface Review and Letters